>>54740723For one of my college assignments 3 years ago, we had to use a Kaggle Pokémon database to practice with making AI classifiers and I chose to try figuring out whether or not a Pokémon is likely to be a legendary depending on certain factors.
After that, I added data from e621 (post count, tags, etc.) to see if there was a correlation between amount of porn/types of fetishes and other things about Pokémon (types, total BST, etc.). I forget the exact results but kek, I should do it again to get more in-depth results. My teacher was shocked, but laughed when I presented it for the whole class.
